Unsafe Sex

Abstract

Unsafe sex, which the Medical Subject Headings thesaurus classifies under sexual behavior, is sexual activity carried out without protection against infection or unintended pregnancy. Cognitive psychology treats it not as a moral failing but as a decision made under risk: a choice in which how the hazard is perceived, how intention is formed, and how affect reshapes judgment in the moment together determine whether protection is used. People systematically underestimate their own vulnerability, form intentions that outrun their behavior, and discount risk sharply when sexually aroused. This article reviews how unsafe sex is explained by risk perception and the optimism bias, by the intention–behavior gap, by affective models of in-the-moment decision making, and by the integrated behavior-change models that draw these strands together.

Keywords: unsafe sex, risk perception, optimism bias, intention-behavior gap, sexual decision making

Introduction

Unsafe sex is sexual activity undertaken without protection against sexually transmitted infection or unintended pregnancy — most often condomless intercourse outside a mutually monogamous, tested partnership. As a form of sexual behavior it is the behavior that health campaigns try hardest to change, and cognitive psychology approaches it with a specific question: given that the hazards are widely known, why do people who can state the risk nonetheless forgo protection? The answer is not ignorance but the structure of the decision itself (Fisher & Fisher, 1992).

That is the thread of this article. Unsafe sex is a judgment under uncertainty, and it inherits every documented quirk of such judgments: the hazard is probabilistic and delayed, the protective act is immediate and costly to the moment, and the person weighing them does so with a mind that discounts distant harms, overrates its own safety, and responds to feeling as much as to fact (Slovic, 1987; Loewenstein et al., 2001). Seen this way, unsafe sex is less a failure to know the odds than a predictable outcome of how human beings estimate, intend, and act on risk.

The Medical Subject Headings thesaurus indexes Unsafe Sex beneath Sexual Behavior in its classification of behavior and behavior mechanisms. That placement is a retrieval convenience, not a theoretical claim; this article uses it only to situate unsafe sex among related behaviors and otherwise treats the decision on its own cognitive terms.

Risk Perception and the Optimism Bias

A decision about protection begins with an estimate of risk, and the estimate is where the first systematic distortion enters. Slovic (1987) established that lay risk perception is not a calculation of probabilities but a judgment shaped by how a hazard feels — how dreaded, how controllable, how familiar it seems. A risk that is voluntary, familiar, and under apparent personal control is judged smaller than its statistics warrant, and condomless sex with a known or trusted partner fits that description exactly: it feels chosen, ordinary, and controllable, so its perceived risk falls below its actual risk.

Laid over this is a bias specific to judgments about oneself. Weinstein (1980) demonstrated unrealistic optimism: asked to compare their own chances of a negative event with those of their peers, most people rate their own risk as below average — a statistical impossibility in aggregate. For sexual health this comparative optimism means a person can accept that unprotected sex is risky in general while believing that they, specifically, are unlikely to be the one infected. Gerrard, Gibbons, and Bushman (1996) meta-analyzed the relation between perceived vulnerability to HIV and precautionary behavior and found it real but modest: feeling at risk does push toward protection, but the link is weak enough that low perceived vulnerability leaves a wide opening for unsafe sex. The optimism bias narrows the very perception that is supposed to drive the protective choice.

The consequence is that accurate general knowledge and inaccurate personal estimates coexist in the same person. Figure 1 lays out the decision pathway this article traces, from perceived risk through intention to behavior, and the first demonstration makes the optimism gap interactive.

The optimism gap in perceived risk

Most people rate their own chance of a bad outcome below a comparable peer’s — a statistical impossibility in aggregate. Set the risk a person assigns to a typical peer, then set how strongly they discount their own case. The gap is the optimism bias.

The Intention–Behavior Gap

Even an accurate sense of risk, and a sincere resolve to protect, does not guarantee the act. The central finding of the social-cognition literature on sexual health is the intention–behavior gap: the correlation between what people intend and what they do is positive but far from perfect. Albarracín, Johnson, Fishbein, and Muellerleile (2001) meta-analyzed the theories of reasoned action and planned behavior as models of condom use and found that attitudes and norms predict intentions well, and intentions predict behavior moderately — but a substantial share of the variance in actual condom use lies beyond intention. People who mean to use protection often do not.

Sheeran, Abraham, and Orbell (1999) meta-analyzed the psychosocial correlates of heterosexual condom use and identified the variables that do the predictive work: communication about condoms with a partner, prior condom use, and the perceived norms of one’s peers carried more weight than general attitudes or knowledge. The gap is not uniform; it narrows when a concrete behavioral history and a specific plan are in place, and widens when the intention is abstract and unsupported. The best-studied bridge across it is the implementation intention — an if–then plan that specifies in advance when, where, and how a resolved act will be carried out. Gollwitzer (1999) showed that such plans reliably convert a goal intention into action by handing its initiation to a pre-chosen situational cue, which is why a settled plan to carry and use a condom outperforms a general resolve to be careful. Webb and Sheeran (2006) made the point experimentally across behaviors: changing intentions produces only a small-to-medium change in behavior, confirming that a sizable gap is the rule and not an artefact of poor measurement.

For unsafe sex the gap is where prevention most often fails. A person leaves a clinic intending to use condoms, forms that intention genuinely, and still has unprotected sex — because the intention was not translated into a plan, a conversation, or an available condom at the moment of opportunity. The second demonstration models the gap directly, letting the reader raise intention and watch behavior lag behind it, and see how a concrete implementation plan tightens the link.

From intention to behavior

A sincere intention to use protection predicts the act only moderately. Raise the strength of the intention and watch the probability of the protective behavior lag behind it. Adding a concrete plan tightens the link.

Affect and the Heat of the Moment

The models so far describe a deliberator weighing costs and benefits in the cold. Sexual decisions are not made in the cold. Loewenstein and colleagues (2001) proposed the risk-as-feelings hypothesis: at the moment of choice, behavior is driven by the emotions felt toward a risk, which can diverge sharply from a cognitive assessment of its probability and magnitude. When feeling and reason point different ways, feeling often wins. Sexual arousal is exactly such a feeling — a visceral state that reweights the costs and benefits of protection in favour of the immediate.

Ariely and Loewenstein (2006) demonstrated this directly. Measuring young men’s judgments both in a neutral state and in a state of sexual arousal, they found that arousal substantially raised self-reported willingness to engage in risky and unprotected sexual activity and to disregard the precautions endorsed when calm. The decision made in the hot state was not the decision planned in the cold one: the same person, differently aroused, returned different answers. This is a visceral influence — a hot–cold empathy gap in which the calm planner systematically underestimates how the aroused self will behave.

The implication reframes the intention–behavior gap. Part of why intentions fail is that they are formed in a cold state by a planner who cannot fully anticipate the preferences of the aroused decider who will actually act. Protection resolved upon in the clinic is discounted in the bedroom, not through weakness of will in any simple sense, but because arousal has shifted the weighting of the very trade-off the intention addressed. The third demonstration makes the shift visible, letting the reader move a state from cold to aroused and watch the acceptance of sexual risk climb.

The heat of the moment

Arousal is a visceral state that reweights the trade-off toward the immediate. Move the state from cold to aroused and watch the accepted level of sexual risk climb above the level the same person planned while calm.

Integrated Behavior-Change Models

No single bias accounts for unsafe sex, and the field’s applied models are deliberately integrative. Fisher and Fisher (1992) proposed the information–motivation–behavioral-skills (IMB) model: protective behavior requires accurate information, the motivation to act on it (comprising attitudes and perceived social support), and the behavioral skills to carry it out — negotiating condom use, obtaining and using condoms correctly. The model’s force is its claim that information alone is inert; it changes behavior only through motivation and skill, which is why knowledge-only campaigns so often fail to move the rate of unsafe sex.

Catania, Kegeles, and Coates (1990) offered the AIDS risk reduction model (ARRM), a staged account in which a person must first label their behavior as risky, then commit to changing it, and finally enact and maintain the change — each stage a separate hurdle at which the process can stall. Labelling is where the optimism bias bites, commitment is where the intention forms, and enactment is where the intention–behavior gap and the heat of the moment intervene. The staged model thus locates the cognitive phenomena of the earlier sections at specific points in a behavioral sequence, and shows why an intervention effective at one stage may leave the others untouched.

Taken together, the integrated models frame unsafe sex as a breakdown that can occur at any of several points — in the estimate of risk, in the formation of intention, in the possession of skills, or in the moment of enactment — and they direct prevention to the specific point at which a given population’s process is failing rather than to knowledge in general.

Worked Example

Consider two young adults, each counselled at a clinic and each leaving with condoms and an intention to use them. The first, assessed a month later, reports consistent use; the second reports unprotected sex on two of three occasions. Both scored identically on a knowledge test and both intended to protect themselves. What separated them? The cognitive models answer in terms of where each person’s process held or broke.

The first person had a concrete plan — condoms kept accessible, a prior conversation with the partner, a settled personal norm — so intention translated into behavior with little gap (Sheeran et al., 1999). The second person held an equally sincere intention but no plan, rated their personal risk as low despite accurate general knowledge (Weinstein, 1980), and made the actual decision in a state of arousal that reweighted the trade-off toward the immediate (Ariely & Loewenstein, 2006). Table 1 lays the two cases side by side across the cognitive variables.

The example makes the article’s point exact. Knowledge and intention, the variables a naive account would privilege, are held constant and explain nothing; the difference lies in the optimism bias, the missing implementation plan, and the heat-of-the-moment state — the cognitive variables the models identify. Unsafe sex here is not a failure to know or even to mean well, but a failure located precisely at the perception of personal risk and the translation of intention into action (Albarracín et al., 2001; Loewenstein et al., 2001).

Discussion

The cognitive psychology of unsafe sex is the analysis of a single decision — whether to use protection — across the stages at which human judgment reliably departs from the normative calculation. At the perception stage, the hazard is underweighted because it feels controllable and familiar (Slovic, 1987) and because people judge their own vulnerability below their peers’ (Weinstein, 1980; Gerrard et al., 1996). At the intention stage, attitudes and norms produce resolve, but resolve predicts action only moderately (Albarracín et al., 2001; Webb & Sheeran, 2006). At the enactment stage, arousal reweights the trade-off toward the immediate (Ariely & Loewenstein, 2006), and the absence of a concrete plan leaves the gap unbridged (Sheeran et al., 1999).

Integrating these shows why unsafe sex is so resistant to information campaigns. Each stage is a separate point of failure, and knowledge addresses only the first; a person can be fully informed and still succumb at perception, intention, or enactment (Fisher & Fisher, 1992; Catania et al., 1990). The affective models add that the decider who acts is not identical to the planner who intended, because a visceral state has intervened between them (Loewenstein et al., 2001). Prevention that works, on this analysis, is prevention targeted at the specific failing stage: correcting the personal-risk estimate, furnishing implementation plans, building negotiation skills, or arranging that protection is decided and available before the hot state arrives.

None of this reduces unsafe sex to a single cause. The point of the cognitive analysis is the opposite: by specifying which stage of the decision is in view and how judgment behaves there, it explains why unsafe sex persists among people who know the risk, intend to protect themselves, and nonetheless do not.

Current Directions

Recent work has sharpened the picture at the enactment stage, where the heat-of-the-moment effect operates. Scott-Sheldon and colleagues (2016) meta-analyzed the experimental literature on alcohol and sexual decision making and found that acute intoxication causally raises the intention to engage in unprotected sex: alcohol is not merely correlated with unsafe sex through a risk-prone personality but actively shifts the in-the-moment decision, a pharmacological parallel to the visceral influence of arousal. The finding gives the affective account a manipulable, preventable input.

Two further strands extend the social and integrative models. Widman and colleagues (2016) meta-analyzed parent–adolescent sexual communication and found it a reliable, if modest, predictor of adolescents’ safer-sex behavior, locating one source of the communication and norm variables that Sheeran and colleagues (1999) had identified as central. And Protogerou, Johnson, and Hagger (2018) built an integrated model of condom use in sub-Saharan African youth, pooling the strongest predictors across studies and confirming that motivation, norms, and behavioral skills outperform knowledge — a direct validation of the IMB framework in a high-incidence setting. The current agenda is thus to identify, for each population, which cognitive stage is the binding constraint, and to target it with an intervention matched to that stage rather than to information alone.

Common Misconceptions

“People have unsafe sex because they do not know the risks.”

Most people who have unsafe sex can state the risks accurately; information alone changes behavior little, because it acts only through motivation and behavioral skills (Fisher & Fisher, 1992).

“If someone truly intends to use protection, they will.”

Intention predicts behavior only moderately; a substantial intention–behavior gap remains, and sincere resolve often fails to translate into the act without a concrete plan (Albarracín et al., 2001; Webb & Sheeran, 2006).

“A calm decision to protect oneself will hold up in the moment.”

Sexual arousal reweights the trade-off toward the immediate, so the aroused decider accepts risks the calm planner rejected — a visceral influence the planner systematically underestimates (Ariely & Loewenstein, 2006).

“Knowing a risk is serious makes people feel personally at risk.”

An optimism bias leads most people to judge their own vulnerability below that of comparable others, so accurate general knowledge coexists with an underestimate of personal risk (Weinstein, 1980; Gerrard et al., 1996).

Glossary

AIDS risk reduction model (ARRM).

A staged account of behavior change in which a person labels their behavior as risky, commits to change, and enacts and maintains it; each stage a point at which the process can stall.

Behavioral skills.

The practical competences protective behavior requires — negotiating condom use, obtaining condoms, using them correctly; in the IMB model, a necessary mediator between information and action.

Comparative optimism.

The tendency to rate one’s own risk of a negative event as lower than a peer’s; aggregated across a group it is a statistical impossibility, and it weakens the motive to protect.

Hot–cold empathy gap.

The failure of a person in a calm (cold) state to predict how they will judge and act in an aroused (hot) state, and vice versa; central to why planned protection fails in the moment.

Implementation intention.

A concrete if–then plan specifying when, where, and how a resolved behavior will be carried out; it tightens the intention–behavior link by pre-deciding the act.

Information–motivation–behavioral-skills (IMB) model.

An integrated account holding that protective behavior requires accurate information, the motivation to act, and the behavioral skills to enact it; information alone is inert.

Intention–behavior gap.

The imperfect correlation between what a person intends and what they do; for condom use the gap is wide, so resolve predicts action only moderately.

Optimism bias.

The systematic tendency to expect better-than-average outcomes for oneself; applied to sexual health, it lowers perceived personal vulnerability below the objective rate.

Perceived vulnerability.

A person’s subjective estimate of their own likelihood of acquiring an infection; a predictor of precaution, though a modest one.

Risk perception.

The subjective judgment of a hazard’s magnitude and likelihood, shaped by qualitative features such as dread, control, and familiarity rather than by probability alone.

Risk-as-feelings.

The hypothesis that behavior under risk is driven by the emotion felt toward the hazard at the moment of choice, which can diverge from a cognitive assessment of its odds.

Theory of planned behavior.

A social-cognition model in which attitudes, subjective norms, and perceived control shape intention, which in turn shapes behavior; a leading framework for condom-use research.

Unsafe sex.

Sexual activity without protection against infection or unintended pregnancy; in cognitive psychology, a decision made under risk rather than a simple failure of knowledge.

Visceral influence.

The reweighting of costs and benefits produced by a hot drive state such as arousal, hunger, or craving, which biases choice toward immediate gratification.

Key Researchers

Dolores Albarracín

is a professor at the University of Pennsylvania whose meta-analysis of the theories of reasoned action and planned behavior as models of condom use quantified how attitudes and norms shape intention, and how far intention falls short of predicting the protective act.

Wikipedia - Wikidata - Google Scholar - Faculty page

Dan Ariely

is the James B. Duke Professor of Psychology and Behavioral Economics at Duke University whose experiments on the heat of the moment showed that sexual arousal substantially raises willingness to engage in risky and unprotected sexual activity relative to the same person’s calm judgments.

Google Scholar - Wikipedia - Wikidata - Faculty page

Jeffrey D. Fisher

is professor emeritus of psychological sciences at the University of Connecticut whose information–motivation–behavioral-skills model established that changing AIDS-risk behavior requires motivation and skills, not information alone.

Wikipedia - Faculty page

Frederick X. Gibbons

is a professor of psychology at the University of Connecticut whose meta-analytic work on perceived vulnerability to HIV established that feeling personally at risk predicts precautionary sexual behavior, though the relation is modest.

Wikipedia - Wikidata - Google Scholar - Faculty page

George Loewenstein

is the Herbert A. Simon University Professor of Economics and Psychology at Carnegie Mellon University whose risk-as-feelings hypothesis explains why behavior under risk is driven by emotion at the moment of choice, often diverging from a cognitive assessment of the odds.

ORCID - Wikipedia - Wikidata - Google Scholar - Faculty page

Paschal Sheeran

is the Stephen J. Walsh Distinguished Professor of Psychology at the University of North Carolina at Chapel Hill whose meta-analyses established both the psychosocial correlates of condom use and the size of the intention–behavior gap that separates resolve from action.

Google Scholar - Faculty page

Paul Slovic

is a professor of psychology at the University of Oregon and president of Decision Research whose foundational work on the perception of risk showed that lay risk judgment is governed by qualitative features of a hazard rather than by probability alone.

ORCID - Wikipedia - Wikidata - Google Scholar - Faculty page

Neil D. Weinstein

is distinguished professor emeritus in the Department of Human Ecology at Rutgers University whose work on unrealistic optimism established that most people judge their own risk of future negative events, including illness, as below that of their peers.

Google Scholar - Faculty page

Laura Widman

is a researcher at Florida State University whose meta-analysis of parent–adolescent sexual communication established it as a reliable predictor of adolescents’ safer-sex behavior, locating a source of the communication variables central to condom use.

ORCID - Wikidata - Google Scholar - Faculty page

Frequently Asked Questions

What is unsafe sex in cognitive psychology?

In research, unsafe sex is sexual activity without protection against infection or unintended pregnancy, treated as a decision made under risk rather than a failure of knowledge. Cognitive psychology studies why people who know the risks still forgo protection, locating the answer in perception, intention, and in-the-moment affect (Fisher & Fisher, 1992).

Why do people have unsafe sex if they know the risks?

Because knowledge is only the first stage of the decision. An optimism bias lowers perceived personal risk, intention translates into action only imperfectly, and arousal reweights the trade-off in the moment, so an informed person can still fail at perception, intention, or enactment (Weinstein, 1980; Albarracín et al., 2001).

What is the optimism bias in sexual health?

It is the tendency to judge one’s own vulnerability to infection as below that of comparable others, which is statistically impossible in aggregate. It lets accurate general knowledge coexist with an underestimate of personal risk, weakening the motive to protect oneself (Weinstein, 1980; Gerrard et al., 1996).

What is the intention–behavior gap?

It is the imperfect link between intending to use protection and actually using it. Meta-analyses find that intention predicts condom use only moderately, so a sincere resolve often fails to produce the act, especially without a concrete plan (Albarracín et al., 2001; Webb & Sheeran, 2006).

How does sexual arousal affect the decision to use protection?

Arousal is a visceral state that raises willingness to take sexual risk. Experiments show that the same person, when aroused, accepts risks and disregards precautions endorsed when calm, so decisions made in the moment differ from those planned in advance (Ariely & Loewenstein, 2006; Loewenstein et al., 2001).

Does alcohol cause unsafe sex?

Experimental evidence indicates it does, not merely correlates with it. A meta-analysis of the experimental literature found that acute intoxication causally raises the intention to engage in unprotected sex, paralleling the in-the-moment influence of arousal (Scott-Sheldon et al., 2016).

Why do information campaigns often fail to reduce unsafe sex?

Because information acts only through motivation and behavioral skills, which it does not supply on its own. Integrated models show that a fully informed person can still lack the motivation, the skills, or the plan to protect themselves (Fisher & Fisher, 1992; Protogerou et al., 2018).

What makes protective intentions more likely to succeed?

Concrete supports: a prior conversation with a partner, a settled personal norm, a history of protected sex, and an implementation plan that pre-decides the act. These narrow the intention–behavior gap more than general attitudes or knowledge (Sheeran et al., 1999; Widman et al., 2016).

References